Skip to main content

Full text of "Smoke control as a part of building fire protection"

See other formats



Summary Minutes of Seminar on May 3, 1977 


National Fire Prevention and Control Administration 

Washington, D.C. 20230 






Central Office Headquarters 

General Services Administration Consumer Product Safety Commission 


Headquarters Headquarters, Drug Enforcement Admin. 

U. S. Department of Energy U. S. Department of Justice 


National Fire Prevention § Control Admin. Headquarters 

U. S. Department of Commerce Tennessee Valley Authority 


Office of the Secretary Headquarters, OSHA/OFASP 

U. S. Dept. of Health, Education, § Welfare U. S. Department of Labor 


Corps of Engineers Headquarters 

U. S. Department of the Army U. S. Dept. of Housing § Urban Development 


Headquarters Chesapeake Division 

National Aeronautics § Space Administration U. S. Department of the Navy 


Central Office Headquarters 

General Services Administration Consumer Product Safety Commission 


1. Highlights . 

The following highlights are emphasized with no order of importance implied: 

a. Test procedures have not been updated to reflect life safety hazards of modem man-made polymers, 
plastics, and synthetics. 

b. Firefighting by building occupants without self-contained breathing apparatus may be deadly. 

c. Exposure to minute quantities of hydrogen chloride (HC1) in air can cause disabling choking and 
burning sensation to eyes, nose, and throat. Polyvinylchloride releases approximately 58% of its 
weight as hydrogen chloride in a fire. 

d. There are a number of factors, other than the heat of the fire, which allow delivery of dangerous 
quantities of smoke and toxic gases to locations remote from the fire regardless of design con- 
siderations, e.g. inversion or lapse, external temperature, wind direction, and building 
operations at the time of the fire. 

e. When a building is heated, the difference between inside and outside temperature causes so-called 
"stack ffect." Due to pressure differences, air is drawn into any ventilating stack (stairways, 
elevator shafts, floor openings) in the lower part of the building. At about the mid-height of 
the building, there is a neutral plane (zone) where air flow is indeterminate. Above the neutral 
plane the flow of air is from the stacks to the floors. Gases from a fire below the neutral plane 
will follow the existing air flow paths and be delivered to floors above the neutral plane. Many 
variables affect the condition in a specific building. When a building is cooled below outside 
temperatures the affect is reversed. 

f . Tests on non-pressurized buildings indicate the following: 

(1) Fire on a floor below the neutral plane can be spread into the corridor by airflow. 

(2) Smoke from floors beneath the neutral plane can be spread to upper floors by the air handling 

(3) Smoke on upper floors will be forced outside. 

(4) Weather can have a significant effect on pressures. 

g. Some tests have indicated that pressurization of corridors to reasonable levels on floors above 
the neutral plane can contain a fire in its room of origin. The use of regular air handling 
systems is practical for most cases of smoke control and minimizes costs . 

h. Pressurization is considered a sound method of smoke control. Further evaluation may lead to 
designs permitting the use of elevators by occupants for evacuation. 

i. At present, there is a lack of complete understanding and knowledge in engineering smoke control 
systems . It is more of an art than a science and adoption of legislation without additional 
knowledge and experience can be counterproductive. 

j . No one can actually state at this time for a given building precisely what amount of pressurizption 
is required to achieve desired results . The only means of determining adequacy is by full-scale 
testing with a trace gas, such as SFg. 

k. Automatic smoke control mechanisms are subject to malfunction. 

1. Wider use of smoke control systems is inhibited by potential liability considerations. 

m. Research at the National Bureau of Standards and the Canadian Building and Research Council 
indicates that six air changes per hour is the maximum needed for smoke control; however, a 
ndnimum of three air changes may possibly be acceptable. 

n. Smoke control requirements will virtually eliminate the need for fire dampers. 

o. Based on locaL- research, Atlanta, Georgia has required that all high-rise buildings built since 
1973 include pressurization systems for smoke control. 

p. In order to achieve reliability in smoke control systems, a design manual is essential. 

Reconnendations . 

a. Federal agencies should utilize research funds to assist in the formulation of a smoke control 
guide manual. 

b. Existing data about smoke control systems must be transformed into charts, graphs, tables, and 
formulae to include all parameters, such as wind, temperatures, building geometry, volume, and 
other special features as a "state-of-the-art" basic reference. 

c. In the design of new and rehabilitated buildings, smoke control should be seriously evaluated 
for feasibility for inclusion. 

d. Pre-fire planning for buildings containing smoke control systems should include contingency plans 
for the various stack-effects which may occur in the building at various times (e.g. various 
seasons of the year) and the need for self-contained breathing apparatus for building occupants. 

Richard Klinker 
January 24, 1978 

Highlights and Recommendations by the NFPCA's 

Federal Facilities Design Standards Task Group i 

Summary Minutes of Seminar ...... 1 

Richard Klinker, General Services Administration 


Robert Taylor, Republic Steel 

Richard Masters, Jaros, Baum & Bolles, Consulting Engrs. 

Francis Fung, National Bureau of Standards 

Brooks Semple, Fire Prevention Consultant 

William Schmidt, Veterans Administration 

Robert Taylor, Republic Steel 



MR. KLINKER: The Task Group on Federal Facilities 
Design Standards is open to all Federal agencies. 
The Task Group is set up to develop standards and 
guidelines for fire protection and fire prevention 
relating to buildings, structures and facilities. 
The areas of concern are occupancy and fire pro- 
tection systems and equipment for assuring ade- 
quate consideration for life safety, continuity 
of operations and property protection. 

The Task Group will assist in reviewing plans for 
new constructions, make independent studies of 
Federal buildings and property and consult with 
appropriate organizations. Thus the Task Group 
serves as a focal point for the various govern- 
ment agencies to learn what other agencies are 

MR. TAYLOR: The concept of the use of air for 
smoke and fire control is at a stage where sound 
design criteria must be available to architects, 
engineers, the fire profession, building code 
officials and others who have jurisdiction over 
the construction, maintenance and control of 
buildings . 

Participants in the meeting include Dick Masters 
of Jaros, Baum & Bolles. Mr. Masters is chairman 
of the Subcommittee on Control of Fire and Smoke. 
Bill Schmidt of the Veterans Administration is the 
current chairman of the NFPA 9QA Conmittee. 
Francis Fung represents the National Bureau of 
Standards Fire Technology and Record Section. He 
is chairman of the Handbook Subcommittee . Brooks 
Semple is Executive Director of the Smoke Control 
Association and secretary of the NFPA 9CA Com- 
mittee. John Fothergill is the last speaker. 

Discussion followed on the National Disaster 
Survival show on network TV. It was felt that the 
show depicted several myths on life and fire 
safety which should be rectified. The National 
Safety Council will be asked to carry another TV 
special program to properly set the record straight. 

The residence fire and high-rise procedures were 
seriously criticized. 

Mr. Taylor then began a slide presentation on 
smoke and toxic fire gases. When a fire occurs 
the code mandates that the air system be turned 
off thus creating a "witches brew" of smoke and 
toxic fire gases from cellulosic products. 

The chance of survival in many buildings is jeo- 
pardized by test procedures which have not caught 
up with the real life-safety hazards of many 
modem materials. Construction practices, too, 
often compound the problem of exit from and access 
to buildings and make smoke control and fire con- 
trol difficult at best. 

Building air systems and certain hazard control 
techniques can dramatically reduce both life and 
property risks. 

The entire fire technology of the recent era was 
based on control of the "wood" fire. During this 
period so-called "fireproof" building construction 
came into being. Codes were aimed basically at 
protecting the integrity of the structure from 
fire. After numerous disasters, life- safety codes 
grew in importance and use. 

Firefighting concentrated on putting out the fire. 
It was the age of the "smoke-eater" with smoke from 
cellulosic materials. Even though wood smoke is 
very high in the release of aldehydes and carbon 
monoxide, few considered toxic gases. 

The concept of "fire load" was born to judge the 
degree of fire hazard of a building. This meant 
the fire endurance of a structure was based on the 
pounds of combustible materials in its construc- 
tion and contents. 

The ASTM E-119 test was developed and the E-119 
typical fire curve is still used to determine the 
degree of fire resistance and performance of a 
system or material. 

There is a new technology of man-made polymers , 
plastics and synthetics. These polymers are com- 
pletely burned in less than five minutes, and the 
rate-of-smoke development is very great. Toxic 
hazards of the fire gases and flame spread can be 
so great that so-called "first aid" firefighting 
without self-contained breathing apparatus may be 

The Ohio State Combustibility Test Method was 
reviewed. This test still uses wood as the material 
to evaluate the hazard of a fuel load. 

The Federal Trade Commission has ruled that many 
of the test methods and standards adapted to 
cellular plastics are "not accurate indicators of 
the performance of such materials under actual 
fire conditions." 

Further, the E-39, Hazards of Materials Committee 
of ASTM, is reviewing all Fire Hazard Test Methods 
and Standards for their relevancy under actual fire 
conditions . 

PVC, for instance, releases approximately 58 percent 
of its weight as HCL or hydrogen chloride. Less 
than 200 parts per million of this gas combined 
with other gases when it decomposes can "arrest" 
an exposed person. The combined effect of these 
gases on an individual in a fire situation will 
depend on the amount of material involved, the rate 
at which gases are released, the size of space into 
which they are released, available oxygen, and the 
ability to evacuate the area. 


Research is being done at the University of Pitts- 
burgh which indicates that more potent sensory 
irritants than HCL may be involved. In a report 
presented at the International Symposium of 
Toxicity and Physiology of Combustion Products, 
it was revealed that HCL has been implicated as a 
major contributor to the overall toxicity exhi- 
bited by the thermal decomposition products of 

The University of Pittsburgh work also indicates 
no difference in the degree of severity of the 
respiratory effects suffered from the thermal 
decomposition products of plasticized (PCV) as 
long as oxygen was above 14 percent, carbon diox- 
ide below 10 percent and carbon monoxide less than 
2000 ppm. 

This data indicates that exposure to 0.10 to 0.22 
milligrams per liter of air of HCL would be 
rapidly incapacitating with choking and burning 
sensation of the eyes, nose and throat. It also 
shows quite clearly that HCL alone is not the 
sole incapacitating agent. 

Empirical gas concentration data and tests indi- 
cate that one pound of PVC, when burned, can 
release concentrations of HCL of 200 ppm into a 
corridor 10x10x50. Thus, a very small fire in- 
volving small surface areas of PVC could have 
large potential hazardous consequences to people 
moving in such a corridor. 

Likewise, Fire Hazard Characteristic 3, "Smoke," 
carries many ramifications. Is it dark or clear, 
light or heavy, hot or cold, filling the entire 
area floor to ceiling or stratified? 

Although gases and many early products of com- 
bustion are colorless, they are components of 
smoke. Most test standards, however, are written 
based on visible smoke, not invisible gases. The 
rate at which smoke is given off by a given piece 
of material is important and determines hazardous- 

The scope of the hazard has been changed with the 
high-rise construction and enormous growth in 
vast area structures. Through design and con- 
struction practices, firefighting is complicated 
in many ways. 

Exterior factors which seriously hamper fire- 
fighting include (1) exterior landscaping which 
makes emergency approach difficult; (2) light 
construction parking ramps prevent apparatus from 
reaching buildings; (3) turns which are too 
narrow for fire apparatus approach; (4) fire 
hydrant placement too far away. 

There is a great need for careful, detailed 
planning of a structure and its surroundings at 
the design stage concerning basic fire detection 
and smoke and fire gas control procedures, alert- 
ing and evacuation programs, and careful plans 
on fighting the fire from the inside. 

The National Research Council of Canada has 
determined that total evacuation is not practical 
from any building over 30 stories in a reasonable 
period of time. Even in a 20-story building, it 

takes up to 20 minutes to evacuate people in 
emergency drills. 

Codes still regulate the service materials and 
exterior finishes for corridors based on fire 
tests not applicable to modem materials. The 
spread of fire can be dramatically reduced by 
requiring all corridor materials to be noncombustible . 

The carpet should meet the new criteria for fire 
safety established by the National Bureau of Stan- 
dards using a radiant heat test tunnel. 

There is another myth that if a building is 
sprinklered, combustible materials may be used 
safely and the corridor length extended beyond 100 
feet between exits. Unfortunately, with modern 
polymers, the smoke load and flame travel rate may 
be so great and move so rapidly before the sprinklers 
operate that a corridor quickly can become non-usable. 

Regarding school fire drills, school corriders 
and classrooms, even though nonconbustible , 
have proved to be hazardous in fire experience 
where basically wood is involved. 

Next is the elevator. The message everywhere is 
"do not use it in case of fire." Elevators by law 
in many areas return to the ground floor. 

The problem of why smoke goes up an elevator shaft 
should be analyzed by those writing the laws. It 
has been shown by work from Brown University and 
the National Bureau of Standards that smoke will 
move up to 50 feet per minute for the first three 
minutes of a fire, and then up to 100 feet per 
minute . 

Vertically, that is 10 stories a minute. The smoke 
can be stopped by closing the flue at the top of the 
elevator. Again, an understanding of the aero- 
dynamics of heat flow in a building can solve many 
of the problems. 

The vent at the top of an elevator shaft is a major 
factor in permitting many working fires to quickly 
involve an entire building. 

Design points which deserve consideration include 
non- fire-stopped shafts, locked exit doors, com- 
bustible ductwork, combustible coatings, combustible 
plumbing materials, etc. 

Attempts are being made to control construction and 
contents with test procedures which often bear little 
relationship to how they burn in a real fire. Little 
has been done to control quantity and rate of smoke 
and gas release of materials. The subject of fire 
technology is not taught in most architectural and 
structural engineering schools. Fire spread should 
be included in the fire reporting. 

Many colleges and universities are now developing 
courses on fire technology. Standards groups are 
evaluating relevant hazards and tests for them. 
The American Society of Heating, Refrigerating and 
Air Conditioning Engineers is developing a new con- 
cept of using air under pressure to reduce the real 
potentials of a high-rise towering inferno. This 
use of air for smoke and fire control is not yet 
an exact science, but the record so far is impressive. 


There is a "neutral point" usually near the middle 
of the building where air pressures are static or 
"neutral." Below this point, fire gases in non- 
pressurized buildings move into the building. 
Above the "neutral zone" smoke and fixe gases 
move out. 

In July 1972 a number of room burnout fire tests 
were conducted at the Henry Grady Hotel in Atlanta 
to determine the effectiveness of stairwell pres- 
surization and the performance of certain 
materials under fire conditions in "low hazard" 
occupancies. The details of these fire tests were 
discussed by Mr. Taylor. The results of these 
tests are available in booklet form from Norman 
Koplon, Architectural Engineer for the City of 

The fact that pressurization works was proven in 
an actual fire on March 7 , 1974, at the Carlyle 
Apartment in Lakewood, Ohio. This apartment had 
corridor pressurization. Everyone evacuated 
safely and the fire was blackened down 15 minutes 
after the alarm was sounded. 

Unusual was the tightness of the corridor and the 
almost complete burnout of the suite in which the 
fire occurred with almost no conmunication to the 

The data gathered from this fire tends to confirm 
work done by the Canadian Building Research 
Council, the National Bureau of Standards and 
others. It shows: (1) fire oh a lower floor 
below the neutral plane can be spread into a 
corridor by the airflow; (2) smoke from floors 
below the neutral plane can be spread to upper 
floors by entrainment into the supply air system 
if the system is not balanced to provide a posi- 
tive supply to all floors under all conditions; 
(3) fire on upper floors can be contained in the 
fire room of origin by the airflow from a posi- 
tively pressurized corridor; (4) smoke on upper 
floors will be forced outside; (5) weather can 
have significant effects on the pressures in the 

From the study of the Carlyle it would appear that: 
(1) Positive pressurized corridors using a "make- 
up" air system can contain a fire within a suite 
if: (a) the system is properly balanced, or (b) 
the fire occurred in a suite above the neutral 
plane. (2) With pressurized corridors the follow- 
ing are essential: (a) stairwells should be pres- 
surized more than corridors to keep smoke and 
fire gases out of stairwells and vents at the top 
of the elevator shafts should be closed; (b) non- 
combustible corridor construction, including doors, 
door frames, door hardware, acoustical ceilings, 
wall coverings and service and conmunications con- 
duit and pipe are critical since radiant heat is a 
factor despite pressurization; (c) corridor 
exhaust systems should be of a non-combustible 
material; (d) suite exhaust systems should be non- 
combustible; (e) all corridor doors should have 
automatic self -closers ; and (f) each fire compart- 
ment should have exterior windows or vents. 

Positive air supply at approximately 0.10 inches 
of water pressure kept this fire out of the corri- 
dor even though the door had failed. Requirements 

for 0.15 inches of water pressure appear excessive 
in condomrniun- apartment type buildings. 

Since that fire, there have been others where corri- 
dor pressurization has worked. Pressurization does 
work, sometimes even without an air system as 
illustrated by a fire in Dallas, Texas. 

The question was raised as to whether or not air 
conditioning systems in high-rise office buildings 
take off early fire gases to the point where fires 
have a hard time developing. 

Pressurization is a sound firefighting tool. With 
it fire m en can get to the scene and fight a fire. 
If life-safety is truly our concern, then smoke 
detection, early voice alerting and pressurization 
are the keys to future life-safety from developing 
fires in fire-rated structures. 

Much has been learned about flashover and the need 
for full-scale room burning. But smoke and toxic 
fire gases in the early stages outside the immediate 
fire area are most critical to human life. 

Through pressurization the elevators can be protected 
and used to move people sensibly. We urge your support 
on the design use of air systems for life-safety 
and property control in buildings. 

MR. MASTERS: Problems which the design engineer 
faces today are numerous. There is a lack of complete 
understanding and knowledge of methodology in engineer- 
ing smoke control systems. 

The kind of problem that design professionals are 
confronted with is illustrated by a statement from 
the latest NFPA 101 Standard— a life-safety code 
which many use in design. "Access corridors in 
buildings seven or more stories in height shall be 
continuously pressurized to a minimum of .01 inches 
on any unit door." 

Conspicuously absent from this particular statement 
is any limitation on the maximum pressure that may 
be developed. If one is designing to a minimum, 
then there must be some factor put into the design 
to make certain that that minimum value is always 

In bfessachusetts, as it relates to the amount of 
air and the amount of pressure that has to be 
maintained in stairwells, the same .01 number 
appears in the latest Conmonwealth of Massachusetts 
Code relating to air pressurization, and requires 
that no less than 60 air changes be provided, 
regardless of building height or configuration of 

Further, it requires that all air be introduced at 
the bottom, regardless of building height. With 
those stipulated code requirements, it may be 
possible to generate a situation which, although 
consistent with the code, could be more dangerous 
than the absence of a pressurization system. 

This points to panic legislation written by people 
who are properly motivated but without the tools to 
either understand or solve the problem. 

The situation exists where people writing the 
legislation do not have the benefit of proper 
backup advice and information relative to limi- 
tations of equipment, dynamics of the forces 
created, fan controls, etc. 

There is a need for some good basic research. 
Good work has been done under the auspices of the 
National Bureau of Standards, through private 
practitioners, various municipal agencies, state 
agencies and Canada. 

Technology has been developed to a point where, 
given a smoke control design task and unlimited 
funds and time, a workable solution could be 
developed for a particular building which relates 
to pressurization and smoke removal. 

However, society is economically oriented and 
people are held personally accountable for their 
actions. The rigors of professional liability 
preclude the undertaking of a design task for 
which there is potential danger to life and the 
acceptance of questionable codes. 

No one can actually state for a given building, 
for a given size and elevation, precisely what 
amount of air will accomplish the desired result. 

If a stairway pressurization has to function 
under all conditions of usage, it must function 
and maintain safe pressures when doors are closed 
and when they are open. It does no good to 
design a system which maintains safe pressures 
only when the doors are closed. Similarly, an 
excessively pressurized stairway will deny access 
to it, and the solution becomes self-defeating. 

An objective should be to develop a simple, 
reliable system that has none or limited automatic 
control. Automatic controls are subject to mal- 
function and should be reduced to an absolute 
minimum in a life-safety system. 

The emphasis of methodology of control in smoke 
control systems has to be looked at quite differ- 
ently. One approach to excess pressure relief 
might be in relieving excess pressure in a stair- 
way by a simple barometric drafting. While this 
arrangement might violate the fire resistant 
integrity of the stair enclosure, a fire damper 
could be added. 

A one and one-half hour fireproof, self-closing 
door in stairways is permitted. When the door is 
open, there is no fire integrity. New thinking 
has to be responsive to the design problem in 
order to relate to relative risks. 

New research is designed to get into the range of 
dynamic systems by fighting fire with the use of 
air systems. A clear definition of the problem 
and an indicated solution must be in the design 
manuals. Engineers need information they can 
work with and information that is proven and 
reliable enough to insure workable design. That 
is what this program is all about. 

MR. TAYLOR: Work done at NBS and the Canadian 
Building and Research Council indicates that six 
air changes per hour is the maximum that is needed. 

It may be possible by good design to get down to 
three or four air changes per hour. Then use of 
the regular air conditioning system is practical 
almost everywhere for smoke control. 

Francis Fung will next speak on smoke control con- 
cepts of different countries. 

MR. FUNG: This slide depicts a smoke situation 
down a 30-foot corridor with both ceiling and walls 
lined with gypsumboard. The carpet is the only 
combustible. The four wood cribs to the end of the 
corridor were the only combustibles being ignited. 
Due to the heat generated from the burn room and 
subsequent heat transfer to the acrylic carpet, in 
only five minutes smoke billows out from the corridor. 
This burning will continue until the carpet is com- 
pletely consumed. 

In a high-rise building, aided by stacking and 
buoyancy, smoke billowing out from a short corridor 
will infiltrate the rest of the building. 

The vertical scale indicates the height of the build- 
ing in feet, and the middle point is the neutral plane. 
The different intensity of the pressure is indicated 
by a relative scale, assuming that the room tempera- 
ture is 70 degrees . A maximum pressure difference of 
.5 inches may be expected. 

The passage of smoke can be devastating in terms 
of life-taking potential, property damage and its 
obstruction to firefighting. 

In order to evaluate the effectiveness of smoke 
systems, a trace gas technique experimental metho- 
dology was developed. This involves the setting up 
of a burn room with warm and pressurized air. This 
pressurized air will simulate smoke from a burn room. 
Infiltration of SF6 tracer gas carried by the smoke 
will be sampled quantitatively to obtain a smoke 
infiltration profile. 

The effectiveness of any given smoke control system 
is compared with the building in normal operations. 
The Seattle Federal Building has a zoned systematic 
pressurized system for smoke control. The concept is 
that the building can be divided into seven zones, 
and given a fire situation, the fire zone in three 
floors will be exhausted. Immediately above and 
below will be pressurized to keep the smoke from 
moving away and infiltrating the rest of the building. 

The stack effect in a stairwell will move smoke which 
is positive at the bottom and negative on the neutral 
plane with respect to the floor. The pressures are 
such in a sheltered interior stairwell that there 
will be less pressure differences than the maximum 
predicted. This is a desirable situation. It means 
in smoke control the building can be designed to 
counter interior stairwell pressure differences 
rather than maximum external differences. 

In a simulated smoke control mode where the middle 
zone around the 20th floor is exhausted and the floors 
above and below are under pressurization, the smoke 
control mode was able to overwhelm the normal pro- 
file and have a pressure profile which will prevent 
the smoke at the middle zone from moving up and below 
the fire zone. 


The San Diego VA Hospital is an interesting build- 
ing with four individual wings, connected by corri- 
dors to a central core. The elevator shafts are 
in the central core. Each floor and wing have 
individual HVAC systems which can be controlled 

Experiments were performed with this building, 
simulating a horizontal smoke control mode. When 
smoke was simulated in the south wing, pressuri- 
zation was exhausted and the west, north and east 
wings were pressurized to keep the smoke from 
moving away from the south wing. In a pressure 
measurement, smoke control mode was achieved. 
When there was smoke control there was no measur- 
able smoke infiltration of the building except in 
the fire wing itself. 

Various smoke control ideas being applied overseas 
in shopping malls are not included in the experi- 
ments from abroad; only those from average type 
high-rise buildings. 

A Canadian experiment recommends that all fire 
areas be pressurized and that the fire floor be 
provided with a means of venting to outdoors by 
smoke shafts or windows. Another suggestion was 
the pressurization of the central core in a high- 
rise building. One other smoke control idea might 
be, in the case of two separate buildings con- 
nected by a bridge on each floor, to separate the 
two buildings to prevent fire from spreading 
from one wing to the other. 

The French emphasis is to supply air from the 
bottom of the corridor floor and to pick it up 
from the ceiling of the corridor. It is hoped 
that stratification exists so that smoke gets to 
the top and cool air is supplied on the bottom 
for safe exiting. 

A smoke control method which the British are 
using consists of distributing air supply into 
the stairwell and thus developing a pressure 
throughout the stairwell of 58 Newton per meter 
square. The fire floor is vented to the outside; 
thus there is a need to exhaust or evacuate the 
hot combustible airs and smokes from the source. 

Incidentally, smoke is not only toxic by itself; 
it also obstructs firefighting and is also com- 
bustible. When a sufficient amount of this com- 
bustible is accumulated in any wide-open area, 
with the right temperature, flashover occurs. 
It is advisable to exhaust as much of the unbumed 
combustibles from a fire as you can. 

It is the recommended practice in firefighting to 
vent combustible mixtures to avoid a flashover. 

To calculate the force on a door of 20 square 
feet, take 50 times the pressure difference in 
terms of inch of water. Only a .2 inch pressure 
maximum can be permitted. In general, in a fire, 
the pressure attenuates very quickly away from 
the fire. Generally accepted opinion is that 
the energy from the system can overcome the 
pressure of a fire except in the immediate vici- 
nity of a large fire. Away from the fire the sys- 
tem can develop sufficient pressure to control 

Smoke control by proper application of HVAC systems 
as a life-safety feature in a large modern structure 
is here today. Many experts around the world strongly 
recomnend it and practice it. However, the various 
smoke control approaches are lacking systematic and 
optimization studies and guidelines. 

The architect does not know what smoke control con- 
cepts are available and which is appropriate for 
his building for the level of protection needed. 
The designer does not have the ready tools to 
transform the concept to HVAC requirements for a 
specific building. 

A systematic and comprehensive study of available 
concepts is needed. Next, a matrix of design guide- 
lines for proper selection of smoke control concepts 
for a given building is required. Finally, optimi- 
zation and design studies have to be performed by 
application of theoretical analysis, large and small 
scale experiments and computer studies. 

This comprehensive study has to be transformed into 
a volume containing charts, graphs, tables and 
formulas — the tools a designer is accustomed to using. 
External effects such as the wind and temperature 
differences have to be included. In addition, building 
characteristics — in terms of geometry, volume and 
special features — have to be considered. 

The manual also should incorporate codes and standards 
that have been enacted on pressure and flow require- 
ments for a specific smoke control. 

This smoke control design manual for engineers is a 
monumental and costly task, but the art of smoke 
control needs this manual for the propagation of 
its cause. Without this comprehensive design tool, 
the progress of smoke control will continue to move 
in an aimless manner. In the end, whatever spark 
or brilliant smoke control ideas that come along 
will remain in the hands of a selected few without 
achieving the full benefit of saving lives. 

MR. TAYLOR: We do not have all the answers on 
pressurization systems. We do know, however, that 
a smoke control system can be designed quite accurately. 
But until the structure is erected and actually tested, 
it is difficult to predict exactly how well it will 
work. This is an area where improvement is needed. 
Smoke control systems are not expensive. Quite the 
contrary. If smoke control design is part of the 
building systems package, the problems and programming 
of the automatic switching of systems can be accom- 
plished off the basic control equipment. 

The problem of wind was one of the design parameters. 
The wind is no problem until the building loses its 

Smoke control systems can be practical and reasonable 
in cost. Available to you today is a tentative research 
project report, TKP 206. 

No one system is going to solve all the problems. 
Early warning detection is still needed as well as 
training of staff. The protective systems already 
in use should be continued. Smoke control is just 
one major new tool which will reduce the potential 
of a fire moving out of the area of origin. 


Brooks Semple will next present a discussion of 
pitfalls of modern design and materials. 

MR. SEMPLE: A very common fire start is one with 
a clean flame, plenty of oxygen and not much 
smoke. This is the way a fire started in an 
automatic transmission plant in Livonia, Michigan. 
This particular fire rewrote the codes for indus- 
trial plants. 

Livonia, Michigan, had a number of acres under 
one roof. A residential bedroom is approximately 
200 square feet. That gathering of smoke at the 
ceiling stops at the side walls and has no place 
to go but down. 

The Livonia, Michigan, fire did not get to the 
stage where the air became contaminated and 
people had to be evacuated. Personnel safety is 
not a major problem in one-story plants, no matter 
how many acres they are . 

Those areas where fire protection engineers have 
been applying their talents are those areas where 
dollars were protected, not lives. 

The first five minutes of a fire are more impor- 
tant than the next five hours. 

As stated earlier, the Henry Grady fire helped 
establish the use of pressurization for smoke con- 
trol. The Atlanta motel test fire further tested 
pressurization, but it was mainly a test of in- 
expensive sprinkler systems for bedrooms. 

The furnishings in the Henry Grady were less than 
normal. There were no draperies nor combustibles 
on the walls or ceiling — only a carpet pad, one 
bed, no overstuffed chair and no papers strewn 
about. This bedroom was furnished more sparsely 
than your own. 

In 3 minutes and 40 seconds, the flame became 
self -extinguishing because the door was closed. 
There is no need to furnish hotel rooms with 
sprinklers. The fires will be self -extinguishing 
as long as the doors remain closed. The people 
inside will be self -extinguishing too. This is 
the new problem we have. 

Built-in extinguishing systems have done a tre- 
mendous job in holding down insurance premiums 
because of the reduction of loss and incidents of 
substantial loss where great masses of property 
are assembled under a single roof. 

When the motel bedroom door was left closed, the 
fire room was vented in 17 minutes following 
self -extinguishment . The fire was reignited, 
and after 18 more minutes the sprinkler opened. 
At that point the fire was no longer visible 
because of the heavy smoke. 

At the moment of opening the door, there was little 
fire pressure pushing the smoke out. In another 
minute the airflow switched around to a natural 
draft. Because fresh air is flowing into the fire 
room, it is clean for the first three feet of 
elevation. Fresh air feeds the fire. It develops 
the fire so that heat builds up and the sprinkler 
head opens. The fire would never have left that 

room of origin and catastrophe would have been 
averted. Sprinkler systems are still the best way 
of averting catastrophe. No where in recorded 
history have 12 or more people been lost in a 
sprinkler building. 

Fusible link manufacturers are given a latitude of 
10 degrees Fahrenheit plus or minus five. A fusible 
link must open between 160 and 170 degrees Fahrenheit 
in water . However , the ' 'UL Standard 33 , Standard For 
Fusible Links" which is also used for fusible links 
on sprinkler heads, allows an ordinary range fusible 
link (from 125 to 170 degrees Fahrenheit) which need 
not open until 290 degrees Fahrenheit. 

On UL's particular time temperature curve, 290 degrees 
Fahrenheit occurs in six and one half minutes. The 
fire may push faster and hotter than that. 

A therma-couple was mounted on the link in the test 
bedroom. The ceiling temperature never exceeded 225 
degrees Fahrenheit until the door was opened. This 
is why the 165-degree links do not open. Thus we 
cannot depend on any heat-sensitive device, as now 
widely marketed, for life safety in the room of fire 

Smoke control depends on smoke detection. Thermal 
devices are too slow. 

Crib bums are familiar to everyone in the fire pro- 
fession. A crib bum was conducted in Sacramento 
four years ago. A small five-room frame house was 
used. The walls were plaster and the wallpaper had 
peeled off. There was no paint on the wall. 

The only combustible part of the house that could 
have been exposed to any fire internally was the 
window and door frames and the carpet padding in the 
living room. 

Because the living room was the largest room in the 
house, it was used for the test. The idea was to 
test the difference between materials commonly used 
in a home a generation ago and those used today. 

Twelve minutes from ignition was self -extinguishing . 
If the infant had been on the floor, it would have 

Today we are in a plastic era. There has been a 
tremendous growth in the use of plastic furnishing. 

In the fire test of the foam mattress and common 
plastic covering on the crib, both cribs were 
identical at 30 seconds into the fire. After that 
period the identity ceases. Flashover occured at 
three and one half minutes. 

We have a substantial problem and need help. 

MR. TAYLOR: A smoke detector, UL labeled, should 
be properly installed in all residences. Secondly, 
be certain that the label is not a ULC label. That 
only identifies conditional certification. Those 
in a position to do so should urge legislation and 
code action. Legislation should be enacted to 
require smoke detectors in all new buildings as 
well as existing structures. 

This new plastic era is here. The wood era is fading. 


In a plastic era, the same time limits do not 
exist. The toxicity is nuch even in a beginning 
fire, that unless easy access is available, the 
fire department should fight the fire. The fire 
department should also be aware of toxic gas haz- 
ards. Even though the fire might seem a clear 
one, gases cannot be seen and masks should be 
worn. It is the clearest smoke which is the worst 
and that occurs in the early stages of fire. 

Bill Schmidt of the Veterans Administration will 
next speak on the design needs of NFPA 90A, the 
Air Conditioning Code. 

MR. SCHMIDT: The NFPA standards are widely used, 
either directly as in the private sector, or 
through various Agency requirements in the Govern- 

Years ago the installation of fire dampers in 
ducts was the way building fire safety was 
designed. About three years ago Appendix B was 
added to the code to permit the systems to be 
used as smoke control systems. 

However, because of the increased awareness of 
smoke and toxic gas danger, much has been done. 
Most of the progress has been in the life-safety 

One section of the code requires that in a build- 
ing with seven or more stories, the corridor be 
continuously pressurized to a minimum of .01 
inches at every living unit door. This means that 
the air conditioning and ventilation systems are 

The 90A code states that the corridor cannot be 
used as a portion of a supply or return air plenum. 
This presents a conflict to the designer. 

The 101 approach is the right way to go. 

Section 9 concerns shopping malls. It requires 
that a covered mall shall be provided with smoke 
control in accordance with 12-37. 

Perhaps the most important section concerns the 
use of smoke dampers and fire dampers in parti- 
tions. Openings in fire partitions for air duct- 
work or air movement shall be protected with fire 
dampers. This requirement need not apply, and 
this is the exception, for ductwork is part 
of an engineered smoke control system. Smoke 
control has almost eliminated the need for fire 

The 90A conmittee has discussed the use of duct- 
work, sprinklers and dampers. 101 has taken fire 
dampers out, which is quite substantial. 

The seriousness of smoke control systems is reflec- 
ted in the life safety code. There is willing- 
ness to trade off fire and smoke dampers to obtain 
a good smoke control system. 

ASHRAE has done a great deal. In addition, there 
is available within the Society reports on build- 
ings which have been built, smoke tested, re- 
designed and retested. 

One of the most important activities in ASHRAE is 
to get material into print and get it referenced 
for use by designers. There is also the insurance 
and professional liability which have to be con- 
sidered. The consensus method must stand up in 
court to protect everyone. 

The NFPA is endeavoring to get as much as possible 
into Appendix B. The building air conditioning 
systems that have been designed and installed for 
normal cooling will be used. This has been found 
to be more than adequate in most cases to do the 

MR. TAYLOR: Many people still Question how much of 
the above is an experiment and how much is proven. 
In Atlanta, every high-rise erected during the past 
four years has had different types of pressurization 
systems designed into it. 

There have been no fires which have gone beyond the 
incident stage in these pressurized buildings in 
the past three years. 

In some of the local governments there has been cost- 
cutting which has forced fire departments and building 
departments into a position of insufficient manpower 
to properly design building fire safety and conduct 
fire prevention inspections along with their fire 
extinguishment duties. 

It will take a national effort to educate local 
officials and the American public that more 
critical than the firefighting may be the building 
review initially and the need for continued inspec- 
tions by both building and fire departments. 

The irony of fire suppression is that by putting out 
the fires and having a low loss record, the city 
councils then cut into the fire departments' budget 
and manpower. This portends a tragedy. 

The air conditioning systems can be used to control 
smoke in the early stages of a fire. The codes should 
be changed to permit the air systems to remain on 
until they are automatically reconfigured to a smoke 
control mode. As long as fresh air is circulating, 
the problems of panic in a fire are dramatically 

The Squibb Building in New York was a case in point. 
The air systems were turned off and 1,600 people on 
upper floors panicked. Windows were smashed. When 
the air systems were turned back on, the level of 
panic in the area dramatically lessened. 

Currently designers do not have a coherent pressuri- 
zation design criteria to work with, as is available 
in almost all other engineering disciplines. In 
addition, designs are based on individual philosophy 
and limited data. There is a need to go in after 
construction and retrotest to be certain that what 
was done was correct. Design guidelines need to 
be printed. Finally, there is the liability which 
exists from unproven tecliniques . 

The actual design technology available is extensive. 
Many of the techniques have been proved and tested. 
However, the technology has not been catalogued for 
what is needed; yet an effective design system is 

dependent upon the structure that it is going into 
and the HVAC system. Critical to this is the HVC 
system and structural leakage. 

What can be done in existing buildings where retro- 
fit cannot be accomplished? Pressurized elevator 
shafts may be one way to stop stack effect in the 
building and create a staging area for firemen. A 
further concept is to then put a fan on the car 
itself to pressurize the lobby area. 

I challenge the concept of taking the elevator car 
out of service during the first three to five 
minutes after an alarm has sounded. There is no 
record of any individual having been injured or 
dying in an elevator in a high-rise building in 
recent years, other than after the HVAC systan 
is turned off. 

Smoke control should be tied into the air systems. 
The system must be economical and effective. 

There is a 10-story hospital which is now divided 
into 30 different smoke zones. People in all but 
three or four of the zones never even know when a 
fire call comes in. This is developing real life- 

Coherent design guidelines can be developed. Limi- 
ted initial effort is underway at the National 
Bureau of Standards, but it must be expanded. 

All existing criteria should be put together in 
one consensus and the existing experimental work 
extended to all significant structural designs. 
This means expanding the National Bureau of Stan- 
dards' work to provide design guidelines for all 
types of structures. This design guide will pro- 
vide the necessary tools needed by engineers, 
architects and building and fire officials to 
judge the validity of the system and to smoke-test 

ASHRAE will endeavor to provide a coherent design 
basis in the form of a manual, as a part of the 
ASHRAE Handbook. The manual will provide an 
effective and rapid dissemination of design tech- 

QUESTION: Where are the sources of information 
now for this type of design? Does ASHRAE have a 
good source list for designing of smoke control 

MR. TAYLOR: ASHRAE has a supply of materials 
available on this subject. NBS also has material 
available on sources. Certain Canadian papers are 
available. However, the data available today is 
fragmentary, and it is difficult for designers and 
architects to pull it out. 

MR. FUNG: The data available at this point is 
basically of a research nature. It is not design 
information in a usable form. There is still much 
work to be done to pool the resources and write 
the design formulas, charts and tables. 

Research is the first step toward technology. But 
there is a great deal that has to be done to trans- 
form this technology into design tools. 

MR. SCHMIDT: The February 1976 issue of the ASHRAE 
Journal was a special fire safety issue. The most 
extensive reference that was available was included 
in the articles. 

QUESTION: Has the engineering problem of laboratory 
buildings where there are multiple hoods with a lot 
of ventilation for toxic materials been addressed? 

MR. FUNG: There are computer programs which will 
simulate this type of building. There are experi- 
ments, comparative programs, small and large scale 
studies which can simulate a building with these 
specific problems. It is costly to run this program 
and the tools are not available to everyone. Still 
needed are charts and formulas to check against each 

MR. SCHMIDT: Actually in laboratory buildings with 
hoods there is a built-in ready exhaust system. 
Let the hoods run. A hood fume will probably deter- 
mine the amount of airflow. The smoke control system 
must have both supply and exhaust. 

MR. FOTHERGILL: The NTH Clinical Center will be 
tested in June and should give some experimental 
data in what is necessary when materials of this 
sort use hoods to maintain evacuation of toxic 
products . 

QUESTION: How can copies of experimental data be 

MR. FOTHERGILL: The best source to contact is 
Francis Fung of the Center for Fire Research at the 
National Bureau of Standards. 

MR. FUNG: ASHRAE and NRC of Canada are the other 
two sources for additional information. 

QUESTION: What effect will energy conservation 
have on smoke control, such as opening windows 
instead of having fixed windows. Does this affect 
the firef ighting operation? 

MR. SCHMIDT: The window problem actually in this 
GSA building is one type of corridor. Most buildings 
are block-buster type with a perimeter of rooms and 
an interior core. 

QUESTION: I was referring to the effect it would 
have on the smoke control system. 

MR. SCHMIDT: The firefighters know whether or not 
they are above or below the neutral plane. An open 
window at the wrong place could either save or kill. 

The entire concept of energy conservation, the load 
shedding of fans at any particular time of day, adds 
an additional complication. Fortunately, most of the 
systems for buildings are now being designed for 

The computerized control system is becoming more 
popular. All the systems are brought in electroni- 
cally with computer surveillance. 

In new buildings the cost factor may be inconsequential. 
So, in some ways the energy conservation has helped. 


VOICE: Do you perceive a trend to go to an HVAC 
smoke control option or elimination of the damper? 
Or do you see some combination of the two? 

MR. SCHMIDT: There will be a caoibination of the 
two because a physical barrier is needed. 

MR. POTHERGTLL: Pressurization is only one tech- 
nuqie. Compartmentalization is a very signifi- 
cant and economical technique. It is closely 
related to both fire protection and energy conser- 
vation. Even a small compartnientalization scheme 
helps in the smoke control area. 

QUESTION: There are many kinds of combustibles 
in areas of these high-rise buildings — in the 
computer room, storage area — where the fire might 
get out of control. During the initial period 
of the fire, there will be additional protection. 
However, in a 2000-degree fire, there are not 
adequate facilities to overcome pressurization. 

MR. SEMPIE: Smoke detectors are extremely impor- 
tant to any smoke control system. The primary 
thing to remember is to call the fire department 
at the first smoke detector signal. 

Smoke control systems are not designed to replace 
fire ratings of walls. They are designed to help 
the fire professionals perform the first two items 
that they must perform. The fireman's first 
function is to ensure that everyone has been 
evacuated from the involved area. 

His second function is to find the fire. With 
the smoke control systems, the fire is being fed. 

QUESTION: What about pressurization systems for 
sprinkler protection? 

MR. SEMPLE: Absolutely not. It is not meant as 
a substitute. Smoke control systems are intended 
to be the first line of life- safety protection 
because they are smoke actuated — thermally actu- 
ated. Extinguishing devices are not fast enough. 

MR. FUNG: Smoke control is only one of the whole 
schemes of fire and smoke safety measures. How- 
ever, it is significant since it concentrates on 
life-safety more so than property protection. 
It is the toxic products that kill, obstruct fire- 
fighting and accumulate combustibles in areas 
away from the fire. 

Smoke control can eliminate the toxic smoke away 
from the origin of fire. It can remove some of 
the combustibles that cause flashover and can 
increase visibility by removing smoke. 

There are concerns about pressure differences due 
to fire. One basic phenomenon of air pressure 
in free expansion is that it attenuates very 
rapidly. In fact, it attenuates as the inverse 
of the distance. 


1. Report No. 


3. Recipient's Accession No. 

I. Title and Subtitle 

MAY 3, 1977 

5. Report Date 

, January 24, 1978 

7. Author(s) R Klinker,R Taylor, F Fung,B Semple,W Schmidt, 
R Masters, W Hanbury 

8. Performing Organization Rept. 


9. Performing Organization Name and Address 

WASH DC 20230 

10. Project/Task/Work Unit No. 


11. Contract/Grant No. 

12. Sponsoring Organization Name and Address 

Jointly sponsored by the Task Committee 5.6 on Fire and 
Smoke Control, American Society for Heating, Refrigeration, 
and Airconditioning Engineers, Inc. (ASHRAE). 

13. Type of Report & Period 


15. Supplementary Notes 

'The meeting was held to discuss the background, development, and status of smoke 
movement and control technology as it pertains to building fire protection. This 
was a jointly sponsored meeting with the Fire and Smoke Control Task Committee of 

In recent years, considerable effort has been devoted to this subject. Other coun- 
tries, as well as the US, have recognized the potential value of using smoke control 
systems to assist in the safe evacuation of buildings while minimizing the adverse 
effects of fires or other emergency situations endangering personnel and property. 
Also, some design choices that have been implemented in new buildings have been 
costly, but accomplished little. 

17. Key Words and Document Analysis. 17a. Descriptors 

Air conditioning 



Fire protection 

Polyvinyl chloride 


Safety Engineering 




17b. Identifiers/Open-Ended Terms 

Combustion products 

High-rise buildings 
Pressurized corridors 
Pressurized stairways 
Smoke control 

17c. COSATI Field/Group 

Smoke movement 
Stack effect 
Toxic gases 

18. Availability Statement 

Release Unlimited. 

19. Security Class (This 


20. Security Class (This 


21. No. of Pages 




USCOMM-DC 8265-P74 


National Fira Prevention and Control Administration 

Washington. O.C. 20230 


Penalty for Private Use. $300 




Third-Class Blk. Rt.